From Mr Simon Briscoe Sir: In many respects the improvements at the Central Statistical Office in the last few years have been impressive but I have a suggestion for Bill McLennan's unfinished agenda ("Lies, damn lies and figures that try to tell the truth", 29 December). The Government's Statistical Service (GSS) still fails to take the initiative. It should take more steps to lead the statistical debate, to educate and to show that it has more expertise than any other organisation. It is only then that the reputation of the statistics can be transformed.

One simple way would be to issue a "statistical brief" at the same time as, say, every third monthly data release. It could include news on improvements or changes to the published data series, reviews of relevant academic work or other research, commen t s on how other countries collect the data and, for that matter, critiques of inaccurate articles in the press or from City analysts. Everyone who takes statistics seriously would benefit from such an exchange.

The unemployment numbers, which are at the centre of this debate, offer a good example of how the initiative could be taken. Mr McLennan proposes publishing figures, defined on the International Labour Office basis, each month; but the production of a monthly count is not on its own sufficient.

The GSS already publishes quarterly data on the internationally accepted definition but the figures get little or no coverage. In part, that is because they are buried in a long release which invariably comes out with other data. If the GSS is serious about shifting the public focus to these "better" figures, why not give them more prominence?

Less prominence should also be given to the monthly claimant-count data, which are widely acknowledged to be hard to interpret. The series does not tell us much and real analysis of the Labour market requires additional data. It is a "lagging" indicator , which has tended to follow long and unexciting cycles - as the chart in your paper showed. Indeed, City analysts probably spend more time explaining why the series is inaccurate than they do explaining the trend! Other countries survive without such a m easure.

I believe that the GSS would build credibility by acknowledging that the claimant count is not worth publishing in all its detail and with the traditional fanfare. Unless people who use, and therefore presumably believe in, the monthly claimant- count d a ta can be found, publication could be in a less detailed form and less timely. On the other hand, if the data do prove to be popular, the critics would be silenced.

Yours sincerely, SIMON BRISCOE Bond Analyst S. G. Warburg Securities London, EC2

28 December